Chapter 8 The Strategies of the Genes Genomic Conflicts, Attachment Theory, and Development of the Social Brain Bernard J Crespi Abstract I describe and evaluate the hypothesis that effects of parent–[.]
Chapter The Strategies of the Genes: Genomic Conflicts, Attachment Theory, and Development of the Social Brain Bernard J Crespi Abstract I describe and evaluate the hypothesis that effects of parent–offspring conflict and genomic imprinting on human neurodevelopment and behavior are central to evolved systems of mother–child attachment The psychological constructs of Bowlby’s attachment theory provide phenomenological descriptions of how attachment orchestrates affective-cognitive development, and patterns of imprinted-gene expression and coexpression provide evidence of epigenetic and evolutionary underpinnings to human growth and neurodevelopment Social-environmental perturbations to the development of normally secure attachment, and alterations to evolved systems of parent–offspring conflict and imprinted-gene effects, are expected to lead to specific forms of maladaptation, manifest in psychiatric conditions affecting social-brain development In particular, underdevelopment of the social brain in autism may be mediated in part by mechanisms that lead to physically enhanced yet psychologically underdeveloped attachment to the mother, and affective-psychotic conditions, such as schizophrenia and depression, may be mediated in part by forms of insecure attachment and by increased relative effects of the maternal brain, both directly from mothers and via imprinted-gene effects in offspring These hypotheses are concordant with findings from epidemiology, attachment theory, psychiatry, and genetic and epigenetic analyses of risk factors for autism and affective-psychotic conditions, they make novel predictions for explaining the causes of psychosis in Prader–Willi syndrome and idiopathic schizophrenia, and they suggest avenues for therapeutic interventions based on normalizing alterations to epigenetic networks and targeting public-health interventions toward reduction of perturbations to the development of secure attachment in early childhood and individuation during adolescence Keywords Attachment Autism Genomic imprinting Schizophrenia Social brain B.J Crespi Department of Biosciences, Simon Fraser University, Burnaby, BC V5A 1S6, Canada e-mail: crespi@sfu.ca A Petronis and J Mill (eds.), Brain, Behavior and Epigenetics, Epigenetics and Human Health, DOI 10.1007/978-3-642-17426-1_8, # Springer-Verlag Berlin Heidelberg 2011 143 144 8.1 B.J Crespi Introduction Conrad Waddington coined the term “epigenetics” to refer to the processes, whereby interactions between genetic and environmental variation lead to the emergence of patterns in phenotypic variation via development (Haig 2004a) In 1957, Waddington explicated in detail his theory of how natural selection and other processes mediate the evolution of ontogenetic trajectories, described by the concept of progressively increasing canalization as depicted in adaptive-developmental landscapes His “strategy of the genes” thus centers on the evolution of cooperative developmental-genetic networks that can, in part via epigenetic modifications, produce relatively viable phenotypes in response to, and in spite of, genetic and environmental perturbations due to alterations in genes or their patterns of expression Epigenetic landscapes represent metaphors for how gene regulation directs pathways of development from the largest to smallest scales – from the first deviation in totipotency of embryonic stem cells, to gene-expression signatures, cell fate determinants and tissue-type differentiation into organs, and to phenotypes that include conditional behavior (Fig 8.1) Much of modern developmental genetics focuses on elucidating the causal processes that underlie subsections of such landscapes, in the contexts of human ontogenies and disease, yet epigenetics itself, the impetus for Waddington’s conception of development, has only recently Fig 8.1 Waddington’s (1957) conceptualization of an epigenetic landscape, with the pegs as genes and the strings as developmental effects from gene expression The Strategies of the Genes: Genomic Conflicts, Attachment Theory 145 reemerged as a central focus of research in understanding the orchestration of phenotypic development For humans, development proceeds via growth and differentiation of body, brain, and behavior under the joint influences of genes and, for environment, mainly interacting humans Throughout much of the most formative early years, the mother holds center stage, providing both physiological nurture from uterus, placenta, and breast, and, increasingly, as the infant grows, psychological guidance through interactive processes of bonding, instruction, and other behavior These developmental processes, perhaps more than many others, are expected to be strongly canalized, yet also sensitive to relatively predictable perturbations, such that epigenetic modifications to gene expression should direct neurodevelopment along conditionally adaptive pathways unless perturbations are too severe As such, and as Waddington (1957) described, epigenetic landscapes may be considered as structures shaped by long evolutionary histories of selection, representing in their form some more or less long-term integration of past social environments and genomes From Waddington’s mid-twentieth century perspective, the adaptive strategies of all genes in an individual coincide, and aspects of the environment, such as the agents of other genomes, are not expected to pursue strategies that conflict Two levels and forms of genetic conflict are now well documented in their effects: conflicts between individuals who are genetically unrelated to some degree, and conflicts within the genomes of individuals, between sets of genes with different patterns of inheritance and genetic relatedness to potential interactants (Hamilton 1964, 1996; Trivers 1974; Haig 1997, 2006a, b; Burt and Trivers 2006; Fox et al 2010) Both forms of conflict lead to differences in the phenotypic optima toward which they are, and have been, directed by natural selection Optimal developmental trajectories of a given individual, and thus epigenetic landscapes, may thus vary in systematic ways between mother and child, and between sets of genes, such as imprinted genes that are silenced when inherited from either the mother or the father (Haig 2002) From evolutionary theory, and now-abundant empirical data (e.g., Hrdy 1999; Maestripieri 2002), offspring are expected to solicit more investment from parents, especially mothers as the main caregivers, than parents have been naturally selected to provide, due ultimately to parent–offspring relatedness of only one-half for autosomal genes Similarly, paternally expressed imprinted genes in the child are expected to exert phenotypic effects that extract more investment from the mother than maternally expressed imprinted genes, due to a history of paternally inherited genes exhibiting lower genetic relatedness within sibships and matrilines (Haig 2002, 2004b) Outcomes of conflict between individuals, and between sets of genes, are difficult to predict, but may include stable equilibria, tugs-or-war over resource allocation, one party “winning” due to asymmetries in control over resource allocation, or continued conflict (e.g., Royle et al 2004; Smiseth et al 2008) Conflicts such as genomic imprinting also potentiate liability to phenotypes associated with disease (Crespi 2010), due to functional haploidy of imprinted genes, dysregulation of tug-of-warbased systems that evolved in this context, and the expected general higher lability of epigenetic gene-expression control systems (based on DNA methylation and histone 146 B.J Crespi modifications) compared to the lability of DNA alterations via mutation Such disease effects from imprinting have been documented extensively for disorders related to human placentation, overall growth, and neurodevelopment (e.g., Angiolini et al 2006; McMinn et al 2006; Wagschal and Feil 2006; Davies et al 2008a, b) The main thesis of this chapter is that genomic conflicts and cooperation, especially the epigenetic effects of genomic imprinting, centrally mediate core aspects of mother–child developmental interactions – most notably the processes of attachment – with important consequences for psychological well-being throughout life I first describe the roles of imprinted genes in development, and recent discoveries of imprinted-gene networks that control growth Next, I explicate the hypothesis that human early-childhood social development, mainly via interaction with the mother, involves a network of brain-expressed imprinted genes that modulate attachment – the process whereby children, in environments characterized by secure and responsive maternal care, internalize psychological constructs derived from external mother–child interactions to develop a self and psyche centered in their social context (Bowlby 1969; Bretherton 1997) The idea of imprinted genes mediating attachment was originally suggested by Isles and Holland (2005), and I extend and evaluate the hypothesis using information from patterns of imprintedgene expression and coexpression, phenotypes of imprinting-based disorders, Bowlby’s attachment theory, and psychiatric conditions involving the social brain Finally, I discuss the implications of these ideas and data for pharmacological and behavioral therapies, public-health strategies, and the integration of epigenetic perspectives derived from Waddington into research programs focused on understanding the genetic bases of human development Most generally, I integrate and synthesize evidence from evolutionary biology, developmental psychology, human genetics and epigenetics, and psychiatry of social-brain disorders to develop and evaluate explicit, testable hypothesis regarding roles of genomic conflicts and epigenetics in human development and evolution 8.2 Genomic Imprinting in Human Growth Haig and Graham (1991) developed the kinship theory of imprinting in the context of conflict interactions in fetal mice between paternally expressed Igf2, which drives growth, and maternally expressed Igf2r, which acts as a “decoy” receptor to reduce Igf2’s growth-stimulating effects Although the predicted pattern of paternally expressed imprinted genes tending to foster overall growth, and maternally expressed imprinted genes constraining it, has been abundantly supported in studies of placentation and body size (e.g., Plagge et al., 2004; Weinstein et al., 2004; Kelsey 2007), the simple, direct tug-of-war system exemplified by Igf2 and its receptor has proven to be atypical of imprinted-gene effects generally The first clear evidence for a much more extensive system of imprinted-gene interaction – imprinted-gene networks – emerged from work by Arima et al (2005), who demonstrated that the imprinted genes ZAC1, LIT1, and CDKN1C jointly mediate The Strategies of the Genes: Genomic Conflicts, Attachment Theory 147 growth of human cells Varrault et al (2006) used information from experimental mouse knockouts, and microarray databases, to document more directly the existence of a coregulatory imprinted-gene network; thus, gains and losses of Zac1 altered the expression levels of the imprinted genes Igf2, H19, Cdkn1c, and Dlk1, and a broad pattern of imprinted-gene coexpression, involving these five genes as well as Grb10, Gtl2, Peg1 (Mest), Sgce, Dcn, Gatm, Gnas, Ndn, and Peg3, emerged from analyses of gene coexpression in pooled datasets from mouse muscle and other tissues Lui et al (2008) subsequently identified a set of 11 imprinted genes in this network, all of which influence aspects of cell proliferation, whose expression levels across multiple tissues paralleled trajectories of overall body growth, and Gabory et al (2009) showed that H19 acts as an important transregulator of this imprintedgene network that may also “fine-tune” gene coexpression patterns to moderate effects from perturbations; network interactions among unrelated genes are indeed postulated as a major cause of robustness against mutations (Wagner 2000), which should be especially important for functionally haploid, imprinted genes Gabory et al (2009) also demonstrated that such regulation via H19 apparently did not operate in the placenta, which implies a notable degree of tissue and stage specificity of imprinted-gene network dynamics Most recently, loss of expression of the ATRX gene in mice has been shown to cause altered postnatal expression of a suite of imprinted genes including Igf2, H19, Dlk1, Zac1, and Peg1, as well as the Rett-syndrome gene MeCP2, suggesting a role for ATRX in transregulation of the imprinted-gene network (Kernohan et al 2010) and corroborating effects of MeCP2 expression in the regulation of imprinted genes (Miyano et al 2008) Two independent lines of data provide further evidence for fundamental roles of imprinted-gene networks in development First, two of the best-understood human genetic syndromes, Beckwith–Wiedemann syndrome and Silver–Russell syndrome, are each mediated by alterations to different imprinted genes in the network, which convergently generate similar phenotypes involving, respectively, overgrowth or undergrowth (Eggermann et al 2008; Eggermann 2009) Similar convergent effects, whereby different epigenetic or genetic alterations produce highly similar phenotypes, have also been described for Prader–Willi syndrome (Crespi 2008a) and Angelman syndrome (Jedele 2007; Crespi 2008a) Convergence from diverse genetic or epigenetic perturbations, to similar phenotypes across multiple traits, represents clear examples of developmental canalization Such canalization effects can also be generalized to idiopathic conditions, such as autism and schizophrenia, each of which also exhibits diverse genetic, epigenetic, and environmental causes converging to a relatively small set of cognitive, affective, and behavioral phenotypes (Happe´ 1994; Owen et al 2007; Abrahams and Geschwind 2008) As imprinted-gene networks have presumably evolved in part due to selection for coordinated, robust control of mammalian growth – yielding specific, syndromic phenotypes when sufficiently perturbed – genomic and epigenetic networks orchestrating human neurodevelopment and behavior may likewise be expected to yield predictable patterns from different forms of alteration A second line of evidence pertinent to gene-expression networks in general, and imprinted genes in particular, is the recent discovery of mechanisms, whereby 148 B.J Crespi imprinted domains interact across different chromosomes, via allele-specific physical juxtaposition of long-range chromatin loops in interphase nuclei (Smits and Kelsey 2006; Zhao et al 2006) Such interactions occur genome-wide (Ling and Hoffman 2007), but are strongly enriched to imprinted regions (Zhao et al 2006), with an apparent central role for the imprinted RNA gene H19 as a hub for transvection of parent-of-origin specific effects to both imprinted and nonimprinted loci on other chromosomes (Sandhu et al 2009) Interchromosomal interactions that involve imprinted loci provide a genome-scale mechanism for coordinated expression of imprinted genes (in addition to mechanisms involving, for example, transcription factors such as Zac1, and protein–protein interactions such as those between p57kip2 and Nurr1) (Joseph et al 2003), and for control of non-imprinted genes and loci by specific imprinted genes, that may serve to increase their relative influence on development In the context of Haig’s kinship theory of imprinting, imprinted-gene networks can be hypothesized as multidimensional generalizations of simple, two-dimensional tugs-of-war, which apparently evolved step by step with the accrual of imprinted domains along the lineages leading to the origins of metatherian and eutherian mammals (Renfree et al 2009) This conception of imprinted-gene networks shows how cooperation, in a literal sense of the word, can evolve from conflict, when the interests of different mutually dependent parties, here paternally expressed and maternally expressed imprinted genes, overlap partially yet broadly Thus, paternal and maternal genes, as well as additional genetic “factions” such as genes on the X chromosome (Haig 2006a), share a common interest in successful development via reducing physiological costs of conflict, and (within limits) increasing the robustness of development to perturbations, although natural selection continues to favor paternal-gene variants that solicit marginally more investment from mothers, and maternal and X-linked alleles that reduce such imposition of increased costs Consideration of the tissue and stage specifics of imprinted-gene expression, in the context of the kinship theory of imprinting, leads to the inference that more or less different imprinted-gene networks should characterize each of the three main arenas for imprinted-gene effects (1) placentation, (2) overall postnatal growth, and (3) behavioral interactions with the mother, as influenced by imprinted-gene expression in the brain Indeed, the network presented in Varrault et al (2006) represents a conglomeration of gene coexpression patterns across multiple tissues and stages, and some central genes in the network, such as Ndn, apparently not exert effects on overall growth (Tsai et al 1999) Eutherian placentation is known to be orchestrated by imprinted genes in a manner consonant with predictions of the kinship theory (e.g., Bressan et al 2009), and alterations to imprinted-gene expression underlie a considerable proportion of risk for the major disorders of human pregnancy (e.g., Charalambous et al 2007); Fauque et al (2010) describe evidence for imprinted-gene coregulation effects in mouse placentation, which coordinate gene-expression changes in relation to early-embryonic conditions Postnatal growth effects are similarly mediated by imprinted genes, which appear to predominantly exert their influences through effects on cell proliferation at the tissue level (Reik et al 2001), and glucose and lipid metabolism at the levels of physiology and metabolism (Sigurdsson The Strategies of the Genes: Genomic Conflicts, Attachment Theory 149 et al 2009) The body growth enhancement effects of Beckwith–Wiedemann syndrome, and growth reduction in Silver–Russell syndrome, indeed represent paradigmatic examples of imprinted-gene disorders with primary effects on both prenatal and postnatal growth Microarray studies that focus on placental gene expression, and gene expression across the tissues most directly involved in prenatal and postnatal growth, are thus expected to reveal imprinted-gene networks that comprise partially overlapping sets of genes, with imprinted genes exerting effects that are more or less tissue specific The tissue with the most pervasive effects on growth, development, and behavior – the brain – remains, however, the least well understood 8.3 Genomic Imprinting in the Brain The study of imprinted-gene effects in brain development was pioneered by Keverne, whose studies of chimeric mice showed differential contribution of paternally expressed imprinted genes to development of limbic brain regions (especially the hypothalamus), and maternally expressed genes to development of the neocortex (Allen et al 1995; Keverne et al 1996; Keverne 2001a, b) Functional and evolutionary hypotheses for the effects of brain-expressed imprinted genes have been described in detail; these include diverse effects on affect, cognition, attention, feeding behavior, and other central brain functions (Isles et al 2006; Wilkinson et al 2007; Davies et al 2008a, b; Champagne et al 2009), some of which apparently influence resource-related interactions between mothers and offspring These hypotheses have proved difficult to evaluate critically due to the complexity of the mechanisms involved, and the difficulty of testing alternative hypotheses of neurobehavioral adaptation compared to pleiotropic by-product Adaptation is most commonly recognized, at least initially, as convergence or parallelism in causal patterns consistent with theory For imprinted brain-expressed gene, clear convergent effects of paternally expressed genes on neurodevelopment have been described for the roles of Peg3, Peg1, and Ndn in promoting development of hypothalamic neurons that secrete oxytocin, the peptide hormone that most strongly mediates social bonding (Davies et al 2008a, b; Ross and Young 2009; MacDonald and MacDonald 2010) Moreover, in the Zac1 network (Varrault et al 2006), these three genes occupy central locations as “hubs”, prominently connected, like Zac1 itself, to relatively large numbers of both imprinted and non-imprinted genes These findings, from three lines of evidence (mouse knockouts, functional gene-expression data, and gene coexpression networks), suggest a functional and evolutionary role for imprinted genes in fostering bonding and attachment of offspring to mothers, in the context of an imprinted-gene network that affects expression of offspring behaviors that regulate levels of resource accrual from mothers In mice, bonding of pups to mothers involves olfactory, tactile, and auditory cues, which foster safety and suckling; moreover, suckling behavior is differentially disrupted in mouse knockouts of the paternally expressed genes Peg3 and Gnasxl (Curley et al 2004; Plagge et al 2004) 150 B.J Crespi In humans, who like mice are highly altricial as neonates, attachment involves the same three categories of sensory cue as in mice, and early suckling and feeding are impaired in the two genomic conditions, Silver–Russell syndrome and Prader– Willi syndrome, that involve strong maternal-gene biases (Blissett et al 2001; Holland et al., 2003; Dudley and Muscatelli 2007), as well as in humans with paternal deletion of GNASxl (Genevie`ve et al 2005); in contrast, macroglossia (enlarged tongues, which are expected to facilitate suckling) has been reported in both Beckwith–Wiedemann syndrome and Angelman syndrome (Cohen 2005) But much more extensively than in mice, human cognitive-affective interactions guide early development of the child’s “social brain” – the distributed, integrated set of neural systems that control acquisition, processing, and deployment of socially interactive information (Frith 2008) Such interactions have motivated the development of a large body of theory and empirical work in psychology and attachment theory, with direct implications for other fields from genetics, epigenetics, and neurodevelopment to analyses of the etiology of psychiatric conditions that involve alterations to early development and function of the social brain 8.4 Attachment Theory and Human Social Development Attachment theory was developed by John Bowlby and Mary Ainsworth to help explain the adaptive significance of physical and psychological interactions between young children and close caregivers (primarily the mother), and how children deprived of care, or subject to dysfunctional forms of caregiver–child interaction, develop along lifelong trajectories characterized by altered emotional and cognitive systems that are explicable in part by the nature of these early perturbations (Bowlby 1969) The majority of children develop “secure” attachment, whereby their intimate interactions with the mother provide for physical safety, nutritional sustenance, and social-emotion-cognitive guidance that generates a “secure base”, increasingly explorative behavior with age, and an environmental conducive to development of a social brain – internalized schema – with robust sense of self, well-developed theory of mind, and ability to nurture secure attachment to one’s own children in later life Deviations from secure attachment, as assayed by the “strange situations test” that tests a child’s behavior toward its mother when subject to short separations, take a small set of forms (1) avoidant attachment, whereby children with unmet expectations of attachment security come to at least provisionally reject significant others, (2) anxious/ambivalent attachment, with unmet solicitation leading to “escalation” of distress and behavior characterized by contact-seeking combined with anger and ambivalence, and (3) disorganized attachment, with lack of a coherent, organized “strategy” for interacting with the caregiver (e.g., Shaver and Mikulincer 2002) These deviations have been interpreted in the context of developmental responses by the child to variation in caregiver sensitivity, that is, provision of consistent physical and psychological support, in contrast to neglectful, hostile, or inconsistent care 8 The Strategies of the Genes: Genomic Conflicts, Attachment Theory 151 The framework for interpretation of childhood attachment patterns has developed in the combined contexts of post-Freudian psychodynamics, evolutionary-ethological constructs for the study of animal behavior, and psychological theories of stages and processes in child social development, such as the social-behavioral internalization theories of Vygotsky (Bretherton 1992; Thompson 2008) In her book Mother Nature, Hrdy (1999) first conceptualized an evolutionary-genetic basis for understanding central aspects of attachment, unappreciated by previous work, that follow directly from W.D Hamilton’s inclusive fitness theory Due ultimately to genetic relatedness between mothers and children of only one-half, mother–child interactions are expected to comprise complex mixtures of cooperation and conflict, with children having been selected to solicit higher levels of behavioral as well as nutritional investment than mothers have been selected to provide The theory underlying parent–offspring conflict has been well supported by empirical studies (Bowlby 1969, p 203; Haig 1993; Hinde and Kilner 2007), and Hrdy (1999) interprets a broad swath of human-specific childhood phenotypes, from high levels of neonatal fat, to precocious neurological development of eye contact, facial expression, and vocalization, as subject to a long history of selection in the context of child signaling of vigor and solicitation of increased energetic and psychological investment Childhood attachment to mother – from placenta, to breast, and to psychological development – is thus expected to be centrally mediated by both cooperation and conflict, which should be expressed in patterns of child attachment behaviors that represent largely adaptive responses to the behavior of the mother, who is in physical control of investment The major patterns of child attachment have indeed been interpreted by attachment theorists as conditionally adaptive responses of children to sensitive, hostile, neglectful, or inconsistent mothering, though not in terms of strategies grounded by the fundamentals of evolutionary genetic and epigenetics Parent–offspring conflict theory applies to autosomal genes with conditional, temporally restricted effects in children, and autosomal genes expressed in mothers The primary implication of this theory is that evolved systems of mother–child interaction (and father–child interaction, given some degree of paternal care) should be characterized in part by conflicts – mainly centered on increased demands from the child, and responses from mother conditioned on marginal benefits to her from incremental investment in child versus benefits from other fitness-accruing activities Haig (1993) describes evidence of such dynamics in the context of human maternal–fetal interactions, as Wells (2003) and Soltis (2004), for suckling and crying; these authors also demonstrate how the major disorders of food provision via placenta and breast, including intrauterine growth restriction, pre-eclampsia, gestational diabetes, excess post-natal weight gain, failure to thrive, and colic, can usefully be understood in part as forms and effects of dysregulation to systems of evolved mother–child conflict – conflicts that are revealed most clearly in cases of genetic or environmental perturbation from “normal” development Mother’s primary form of investment in children beyond food is, of course, psychological training and guidance, and mother–child conflicts in this arena are expected to exhibit constrained-conflict dynamics squarely manifest in attachment, 152 B.J Crespi with disorders of mental health, for the grown child, more likely to result when dynamics are perturbed (Berry et al 2007, 2008; Lyons-Ruth 2008) To understand the expected role of imprinted genes in attachment, conflicts due to imprinted genes must be conceptually distinguished from, and related to, conflicts between parents and offspring Imprinted-gene conflicts are thus expected, from the kinship theory of imprinting, to involve interactions between paternally expressed genes and maternally expressed genes in the child As such, imprinted genes are predicted, through the integration of their effects, to influence the child’s “set level” of resource demand imposed upon the mother An example of such an economic system, in a physiological situation, is provided by expression of the imprinted gene Cdnk1c in fetal mouse development: levels of its protein product, p57kip2, have been experimentally demonstrated to act as a “rheostat” for embryonic growth, with ultimate levels of growth determined by the balance between fetal demand and maternal supply (Andrews et al 2007) In the context of child psychology and behavior, imprinted genes are expected to similarly mediate levels of imposed demand, here for psychological time and energy; more generally, they should exert effects that result, by any mechanism, in higher or lower, longer or shorter, levels of investment from the mother Small alterations to such systems, toward, for example, stronger effects from paternal gene expression that lead to higher investment, are expected to benefit the child, and his or her paternally expressed imprinted genes, at a small cost to the mother and to the child’s maternally expressed imprinted genes In contrast, large alterations, which like any large genetic or epigenetic effects surmount the homeostatic effects of canalized development, are expected to result in disorders of psychological development that are detrimental to both mother and child, and to all genomic parties concerned The nature of such disorders, however, provides useful tests of evolutionary theory involving genomically based conflicts, and insights into strategies for therapy, prevention, and research (Crespi 2008b; Crespi et al 2009) These considerations from attachment theory, parent–offspring conflict theory, and the kinship theory of imprinting generate a simple conceptual framework for understanding normal and dysregulated social-behavioral interactions between mothers and small children (Fig 8.2) From birth, young children exhibit some level of social-behavioral “demand”, manifest in solicitations of interaction via facial and auditory cues, as well as in the contexts of crying, fussing, and suckling Such levels are expected to vary between children, and over time in response to levels and forms of reciprocal and unsolicited maternal behavior Child-specific, dynamic “demand” is matched, less or more, by levels of maternal sensitivity and responsiveness to such cues – varying from neglect, to solicitous, attentive care, and to controlling over-involvement Evolutionary histories of parent–offspring conflict, and imprinted-gene conflict, have generated conditional behavioral adaptations, in child and in mother, that potentiate the possibility of mismatches between demand and response, which may result in maladaptation for one or both parties Thus, relatively low social-behavioral demand from the child may delay or dysregulate social-brain development, but provide marginal benefits to the mother in terms of her lifetime inclusive fitness 8 The Strategies of the Genes: Genomic Conflicts, Attachment Theory 153 Fig 8.2 A simple model for potential interactions of levels of maternal social-behavioral provision/imposition with levels of child social-behavioral solicitation/imposition In contrast, relatively high demands may, if matched by supply, provide benefits to the child at some cost to mother; if not matched or if too insistent, however, such demands may provoke neglect, ambivalence, or even hostility, if they interfere sufficiently with a given mother’s perceived optimal life-history trajectory Levels of maternal “supply” higher than child demand – “over-involvement” – may, in contrast, come to interfere with the child’s ability to explore and increasingly detach, physically and social-emotionally, from its secure base of the controlling maternal brain, to individuate its own psyche This conceptual framework can help to explain the surprisingly low rates of secure attachment in children (not much over 50%), with anxious attachment representing, in part, attempts to solicit increased investment and improve one’s lot, whereas avoidant attachment may involve a form of defensive strategy to, in part, avoid making matters any worse The outcomes of parent–offspring and imprinting-mediated conflicts are difficult to predict a priori, but they should depend upon the relative fitness costs and benefits to the child and mother of the alternative behaviors, and the degree to which each party controls aspects of the interaction Thus, tactics available to young children include auditory and facial expressions of positive and negative affect, crying, suckling intensity, and physical movement varying with age; in contrast, mothers control access to breast and physical presence, and provision of information and social interaction crucial to the child’s affective and cognitive development Indeed, as Bowlby (1951, p 53) noted: “In dealing here with the embryology of the human mind one is struck by a similarity with the embryological development of the human body, during the course of which 154 B.J Crespi undifferentiated tissues respond to the influence of chemical organizers If growth is to proceed smoothly, the tissues must be exposed to the influence of the appropriate organizer at certain critical periods In the same way, if mental development is to proceed smoothly, it would appear necessary for the undifferentiated psyche to be exposed during certain critical periods to the influence of the psychic organizer – the mother” Thus, during infancy and early childhood, “she is his ego and his super-ego.” The primary usefulness of a Keynesian, supply–demand model for attachment is that it provides a framework, compatible with both developmental psychology and evolutionary-genetic theory, that generates novel explanations and predictions for alternative patterns of social-brain development The model can be conceptualized as a developmental landscape, canalized to secure attachment and its sequelae but with different optimal trajectories for mother and child, and for child’s paternal and maternal imprinted genes Child development along the landscape initially proceeds via basic emotional-behavioral attachment mediated in part by paternal imprintedgene effects, which serve as an affective-attentional scaffold upon which the social, maternal brain is gradually built, via maternal-gene effects and interactions with the mother and other caregivers Actual trajectories for any child are determined by some dynamic integration of genetic, epigenetic, and environmental inputs into social-brain development, and into mother–child interaction; most importantly for mental health, relatively large genetic, epigenetic, or environmental alterations to this system are expected to generate predictable consequences, manifest in major disorders of the social brain, such as schizophrenia, depression, and autism The assumptions and predictions of the model can be evaluated using information from human imprinted-gene syndromes, and studies on the etiology of schizophrenia and autism, in the contexts of attachment and evolutionary-genetic theory 8.5 Prader–Willi Syndrome Haig and Wharton (2003) describe how the complacent nature of Prader–Willi infants (with low levels of crying and wakefulness), and their compulsive selfforaging for food after the usual age of weaning, can be interpreted in the context of reduced demands on mothers, as expected under imprinted-gene biases involving loss of paternal-gene expression Prader–Willi syndrome is known to centrally involve hypothalamic alterations and relatively low levels of oxytocin (Swaab et al 1995; Holland et al 2003; H€ oybye 2004; Goldstone 2006), which is consistent with the oxytocin-reducing effects of Ndn gene knockouts in mice (Davies et al 2008a, b) and the expected behavioral effects of lower oxytocin levels – reduced motivation toward social bonding behavior These considerations suggest that Prader–Willi syndrome should involve insecure attachment in children, especially among cases due to maternal uniparental disomy in which the genomic deviations toward maternal imprinted-gene effects are more extensive This suite of behavioral and other alterations in Prader–Willi syndrome may involve, in part, disruptions to genes that underlie a human-lineage trajectory toward earlier weaning and shorter The Strategies of the Genes: Genomic Conflicts, Attachment Theory 155 interbirth intervals (Sellen 2007; Haig 2009; Humphrey 2009), a transition expected to impose strong selection in the context of maternal-offspring and imprinted-gene conflicts given generally high rates of early-childhood mortality The extensions of attachment theory described here provide a simple, novel hypothesis to help explain one of the other major features of Prader–Willi syndrome: the extremely high levels of affective psychosis upon reaching adolescence (Soni et al 2008; Webb et al 2008) By this hypothesis, Prader–Willi infants and children may essentially undergo a process of self-induced maternal deprivation, whereby their low levels of care solicitation lead to reduced interaction with the mother and dysregulated social-emotional development The primary line of evidence supporting this hypothesis, in addition to the theory and data described above, is that maternal deprivation, due to diverse causes including reduced physical care, temporary early separation, or permanent loss, is a well-documented, highly penetrant causal factor in the etiologies of depression (e.g., Pesonen et al 2007; Tyrka et al 2008; Liu et al 2009), schizophrenia (reviewed in Read and Gumley 2008), and schizotypy (Anglin et al 2008) Convergent evidence for interactions of maternal deprivation with imprinted-gene expression comes from a study of mice, where (5.4%) of 37 genes significantly upregulated in the hypothalamic paraventricular nucleus of young mice in response to maternal deprivation were imprinted (Gtl2 and Ipw), a higher proportion than expected given an approximate 1.2% (120 of 10,000) (http://igc.otago.ac.nz/ table.html) of brain-expressed genes imprinted in the mouse genome (P < 0.05) To the extent that affective psychosis in Prader–Willi syndrome is mediated by gene–environment interaction in this context, alterations to early maternal-social environments, and facilitation of early bonding via behavioral and hormonal therapies, may help to alleviate the psychiatric sequelae of this large-scale epigenetic disruption Similar considerations should apply to genetic and epigenetic alterations that yield phenotypes overlapping with Prader–Willi syndrome (Crespi 2008a) and, perhaps, to effects from reduced size of the paraventricular nucleus in patients with bipolar disorder and major depression, compared to controls (Manaye et al 2005) More generally, an evolutionary trajectory toward earlier weaning and shorter interbirth intervals in humans might be expected to have increased vulnerability to affective-psychotic conditions, unless attachment through alloparents and fathers (Hrdy 2009) compensates for abbreviated maternal care 8.6 Angelman Syndrome In contrast to children with Prader–Willi syndrome, children with Angelman syndrome are highly active, with reduced duration of sleep, frequent suckling attempts, and enhanced levels of socially directed positive affect (Cohen et al 2005; Oliver et al 2007) These diverse phenotypes have been interpreted as reflecting increased behavioral and energetic demands on the mother, as predicted by the direction of imprinted-gene deviation, toward reduced relative effects from the maternally expressed imprinted gene UBE3A (Oliver et al 2007) Such an 156 B.J Crespi increased “set level” of demands imposed on mothers might be expected to lead to insecure forms of attachment, unless the demands are met, whereupon socialemotional development should proceed as normally as possible under the circumstances of severe intellectual disability that characterize this syndrome A high prevalence of autistic traits in Angelman syndrome (Sahoo et al 2006; Bonati et al 2007; Jedele 2007) has also been interpreted in the context of selectively reduced social-brain development and high demands imposed on the mother in the context of paternal imprinted-gene bias (Brown and Consedine 2004; Isles et al 2006; Crespi and Badcock 2008) 8.7 Schizophrenia and Attachment Schizophrenia represents a set of conditions, due to diverse causes, with some combination of symptoms that may include hallucinations, delusions, thought disorder, flat affect, and anhedonia (Read et al 2004a, b) An attachment-theory perspective on schizophrenia was first provided by Bowlby (1973, pages 174–177, 318–319), in his description of continuities between forms of abusive and dysfunctional psychological treatment of children by parents and the later development of specific schizophrenic symptoms Systematic analyses of attachment in relation to aspects of schizotypy and schizophrenia (Berry et al 2007, 2008; Gumley et al 2008; Tiliopoulos and Goodall 2009) have documented higher than normal levels of insecure attachment in these conditions, with general support for Bowlby’s prediction of linkages between childhood experiences and psychotic-behavioral profiles Thus, in nonclinical populations avoidant attachment has been associated with negative schizoptypal traits, and anxious attachment with positive schizotypy (Wilson and Costanzo 1996; Berry et al 2006); insecure attachment has also been linked with paranoia in schizophrenia (Pickering et al 2008), and disorganized attachment may be related to dissociative cognitive states that accompany schizophrenia in some patients (Berry et al 2008; Liotti and Gumley 2008) One of the best-replicated patterns reported in families with an adolescent or young-adult schizophrenic offspring is high levels of “expressed emotion” – some mix of parental hostility, criticism, and over-involvement (Read et al 2004a, b), which commonly occurs in the context of emotional over-involvement by one or both parents but low levels of actual care Higher expressed-emotion predicts higher rates of relapse after a first psychotic episode and also appears to be linked with the initial development of schizophrenic symptoms themselves (Read and Gumley 2008; Read et al 2004a, b) The causal bases for such associations have puzzled psychiatrists for many years, but perspectives from attachment theory and genomic conflict offer novel, testable insights Thus, parental over-involvement might be expected to involve attempted control over offspring behavior, interfering with the usual processes of physical and emotional separation of the child from attachment figures, which becomes most pronounced in adolescence (Harrop and Trower 2001) Moreover, extended controlling influences on the offspring psyche The Strategies of the Genes: Genomic Conflicts, Attachment Theory 157 by the mother – the “maternal brain” of genomic imprinting theory (Badcock 2009, pages 154–157) – as expressed in the mother’s behavior and in maternal-gene influences on development within her child – are expected to engender notably higher risk of schizophrenia Such a prediction is broadly compatible with maternal imprinted-gene biases, and their correlates, being differentially associated with higher schizophrenia risk (Crespi 2008a; Crespi and Badcock 2008), but additional, targeted tests are required for robust evaluation More generally, evolutionary theory provides clear, specific predictions regarding the presence and causes of conflicts within families (Emlen 1995, 1997; Surbey 1998), which underpin and complement psychodynamic and socioeconomic perspectives on such conflicts, as well as providing a basis for understanding conflicts within the minds of normal “individuals”, and those beset with such conditions as schizophrenia or depression (Read et al 2004a, b; Haig 2006b; Badcock 2009) As Laing and Esterson (1970, p 12) inquired, “are the experiences and behaviour that psychiatrists take as symptoms and signs of schizophrenia more socially intelligible than has come to be supposed?” Long-term prospective studies, using high-risk populations, are required to evaluate patterns of cause and effect in relating childhood experiences and attachment patterns to the risks and forms of depression, schizophrenia, and other conditions Evolutionary theory provides a conceptual framework to guide such studies, a framework that is fully compatible with both a psychoanalytic focus on childhood social-brain development in the family context and a strong role for interactive genetic and epigenetic effects in modulation of risk In this context, the mother represents the primary environment of the child from conception until well past weaning, an environment that intimately modifies gene expression, and neurodevelopment, in the growing child As such, supportive public-health interventions during early child development and in adolescence, which are sensitive to the main causes of conflicts within families, are expected to yield disproportionate returns in reducing the incidence, severity, and recurrence of schizophrenia and depression, and more generally increase emotional well-being in society (e.g., Gumley et al 2008) 8.8 Autism and Attachment Idiopathic autism represents a spectrum of related conditions, with many genetic, epigenetic and environmental causes, characterized by differential underdevelopment of the social brain from earliest infancy (Kanner 1943; Happe´ 1994; Abrahams and Geschwind 2008) A central focus in attachment-theory research on autism has been whether or not autistic children can indeed develop secure attachment with the mother, given that Kanner’s (1943) original description of autism posited lack or disturbance of affective contact as a central feature Rutgers et al (2004) reviewed 16 studies of attachment in autistic children, finding that autism was associated with lower levels of secure attachment overall, but that this 158 B.J Crespi difference “disappeared in samples of children with higher mental development” or less severe symptoms, suggesting that attachment security is compatible with at least relatively high-functioning autism In contrast, van Ijzendoorn et al (2007) and Rutter et al (2009) question the applicability and validity of attachment theory itself to autism, given the unusual nature of social relationships in this condition and uncertainties concerning whether autistics can develop internal working models of self and others, which form the bases for linguistic discourse Psychological perspectives based in Vygotsky’s work indicate, for example, that linguistic interactions with mother mediate childhood progress from listening, to private speech, and eventually to inner speech as thoughts (Fernyhough 1996), a process that is absent or underdeveloped in autism but may by contrast, be dysregulated via “reexpansion” in the auditory hallucinations of schizophrenia (Fernyhough 2004) A general view of autism posits that under-development of the social brain should, as described by Kanner, centrally involve weak social-emotional connections with other individuals Such differential social deficits, however, need not preclude relatively basic forms of attachment to the mother, as described by Bowlby for mother–offspring relations across most mammals Indeed, from the perspective of parent–offspring conflicts and genomic imprinting, some subset of autism cases might also be considered to involve pathological overexpression of increased levels of demand imposed on mothers by offspring (Badcock and Crespi 2006; Crespi and Badcock 2008) Such demand may be imposed either directly, through behaviors that solicit forms of investment as in Angelman syndrome, or indirectly, through self-oriented, nonsocial behavior that precludes or delays physical independence, requiring mothers or others to provide longer-term, more highly intensive care This hypothesis is concordant with several lines of evidence, including (1) accelerated brain and body growth in young children with autism and increased relative effects from paternally expressed imprinted genes (Crespi and Badcock 2008), (2) imprinted-gene effects on reaction to novelty, and dispersal, in mice (Isles et al 2002; Plagge et al 2005), (3) a higher incidence of autism in males, who are more costly than females to rear (Gibson and Mace 2003; Rickard et al 2007; Tamimi et al 2003), (4) myriad reports of close and sustained, if atypical, relationships between mothers and their autistic children (e.g., Hoffman et al 2009), and (5) temperaments of autistic children that involve higher rates of activity, impulsivity, and noncompliance (e.g., Garon et al 2009) The degree to which autism spectrum traits in nonclinical populations actually engender higher levels of early maternal or biparental investment has yet to be investigated, and indeed, many childhood social abilities are expected to have evolved in the context of soliciting higher investment from mother via providing her with positive-affect emotional benefits contingent on social skill development (e.g., Hrdy 1999), in addition to imposing negative-affect costs through behaviors such as refusal, tantrums, and, perhaps, insistence on sameness For analyzing behavioral interactions between mothers and offspring, it is crucial to bear in mind that “disorders” such as autism are postulated to represent sequelae to dysregulation of mechanisms that evolved in contexts of both conflict and cooperation, with evolutionary and behavioral conflicts potentiating increased levels of risk (Crespi 2010) 8 The Strategies of the Genes: Genomic Conflicts, Attachment Theory 159 A primary implication of attachment theory, and behavioral-evolutionary genetics, for elucidating the causes of autism, is that early-childhood social motivation, in the context of “demands” imposed on mothers via physical, social-emotional, and linguistic interactions, should drive cascading effects on social-brain development that are either normal or deviate toward autistic phenotypes The neurological and endocrine bases for social motivation and affective bonding in infants (e.g., Bowlby 1969 p 203–204; Moriceau and Sullivan 2005; Grossmann et al 2008) and children (e.g., Bartz and Hollander 2008) have been much less studied than those in mothers (e.g., Strathearn et al 2009), though their relevance for understanding mother–child interactions should be no less consequential Studies of plasma oxytocin levels in autism have yielded unusual results: two studies found lower plasma oxytocin in children with autism (Modahl et al 1998; Green et al 2001), but unexpectedly, oxytocin levels were positively associated with degree of social impairment (Modahl et al 1998); in the single study of adults, autistic individuals showed significantly higher levels of plasma oxytocin than did controls (Jansen et al 2006) Might oxytocin in autistics subserve, to some degree, positive reinforcement of nonsocial stimuli, or might epigenetically based reduction of oxytocin receptor expression (Gregory et al 2009) follow from early social-interaction deficits? The m-opioid system represents another important candidate mechanism, in addition to oxytocin, for neuroendocrine regulation of attachment, given links of allelic variants in the m-opioid receptor gene OPRM1 with attachment patterns, sensitivity to social rejection, drug dependence, and risk of schizophrenia (Insel 2003; Barr et al 2008; Way et al 2009; Mague and Blendy 2010; Sery´ et al 2010) and evidence for parent-of-origin effects influencing phenotypic effects of this gene (Lemire 2005) 8.9 An Imprinted-Gene Network for Attachment of the Social Brain Three lines of evidence (1) demonstrations that imprinted-gene expression and coexpression centrally mediate prenatal and postnatal growth, (2) diverse forms of data on imprinted-gene expression effects in the brain, and (3) expectations from theory that mother–offspring interactions should involve genomic conflicts, lead to the prediction that a brain-specific imprinted-gene coexpression network should exist, and should modulate aspects of human mother–child attachment The presence of such a network in humans was evaluated by testing statistically for differentially frequent coexpression of brain-expressed imprinted genes, compared to expression of imprinted genes with other genes independent of imprinting status For each of the 64 verified human imprinted genes (http://igc.otago.ac.nz/table html), a set of highly coexpressed genes were generated using the Gemma coexpression database (http://www.chibi.ubc.ca/Gemma/searchCoexpression.html) These sets of coexpressed genes were combined to generate a list of genes that were highly coexpressed with four or more imprinted genes The list comprised 188 genes in all, of which approximately 0.6% (64/~10,000 total brain-expressed genes) 160 B.J Crespi should be imprinted under a null model of coexpression random with regard to imprinting status Six (3.1%) of the 188 coexpressed genes were imprinted, an approximate fivefold excess (P < 0.001) The highly coexpressed imprinted genes included PEG3 and NDN, two genes that mediate development of oxytocin-secreting neurons (Davies et al 2008a, b); NDN may also be involved in the risk of schizophrenia (Le-Niculescu et al 2007), and both PEG3 and NDN are members of a network of genes differentially coexpressed in the prefrontal cortex of schizophrenics (module 16 of Torkamani et al 2010) Additional relatively highly coexpressed genes in the network included three genes with genetic-association or copy-number variation links to both autism and schizophrenia, CNTNAP2, NRXN1, and BOLA2 (reviewed in Crespi et al 2010; coexpressed with 6, 7, and 11 imprinted genes, respectively) Additional data implicating this network, and imprinting effects, in autism and schizophrenia include (1) the presence of parentof-origin effects for the associations of CNTNAP2 with autism (Arking et al 2008) and for the speech-language associated gene FOXP2 (Feuk et al 2006) that interacts directly with CNTNAP2 (Vernes et al 2008), and (2) colocalization of the imprinted, schizophrenia-associated gene LRRTM1 (Francks et al 2007) with NRXN1 at glutamatergic synapses (Brose 2009) This analysis requires replication using other datasets, and more detailed bioinformatic dissection of coexpression patterns, but it provides preliminary evidence of interactions between brain-expressed imprinted genes that mediate aspects of mother–offspring interaction and social-brain disorders The existence and structure of brain-specific imprinted-gene network should have important implications for pharmacological interventions; for example, of the 11 genes most highly altered in expression from treatment of mice with the mood-stabilizer drug valproic acid (Chetcuti et al 2006), two are imprinted (Peg3 and Sfmbt2), one is predicted to be imprinted (Zic1)(Luedi et al 2007), one interacts directly with the imprinted gene Wt1 (Par-4)(Richard et al 2001), and one (Kcna1) interacts directly with Cntnap2 (Strauss et al 2006) Such a striking concentration of imprinted-gene related expression changes is consistent with the molecular function of valproic acid as a histone deacetylase inhibitor, an agent that exhibits differential epigenetic effects on imprinted genes (e.g., Baqir and Smith 2006); moreover, valproic acid during human or rodent pregnancy is a highly penetrant cause of autism in offspring (Rinaldi et al 2008; Dufour-Rainfray et al 2010), and valproic acid partially restores levels of MeCP2 in a mouse model of Rett syndrome (Vecsler et al 2010) Might such epigenetic treatments exert their effects though canalizing or decanalizing neurological-pathway functions of interacting imprinted genes? 8.10 Conclusions Understanding human social development requires integration of theory and data from diverse, highly specialized, disciplines, including genetics, development, neuroscience, psychology, and psychiatry Conrad Waddington’s still-nascent The Strategies of the Genes: Genomic Conflicts, Attachment Theory 161 science of epigenetics serves as a conceptual and experimental platform to connect social-environmental variation with the gene-expression patterns that drive development, as genes and family, especially the mother, jointly sculpt developing brains Like other traits, epigenetic landscapes of social development have evolved, subject to the conflicts and confluences of genetic interest that form the cornerstone of modern evolutionary theory Effects of parent–offspring conflicts, and the intragenomic conflicts of imprinting, are predicted from such theory to modulate normal development and to potentiate specific directions and forms of maladaptive trajectory, due to genetic, epigenetic, and social-environmental perturbations As such, dovetailing of predictions from evolutionary genetics and epigenetics with the conceptual constructs and experimental tools of developmental psychology, neuroscience, and epigenetics should yield novel insights into the evolutionary underpinnings of the human social brain and its disorders, and the strategies of the genes involved Acknowledgments I am grateful to Christopher Badcock, David Haig, Sarah Hrdy, and Felicity Larson for comments, and Paul Pavlidis for advice References Abrahams BS, Geschwind DH (2008) Advances in autism genetics: on the threshold of a new neurobiology Nat Rev Genet 9:341–355 Allen ND, Logan K, Lally G et al (1995) Distribution of parthenogenetic cells in the mouse brain and their influence on brain development and behavior Proc Natl Acad Sci USA 92:10782–10786 Andrews SC, Wood MD, Tunster SJ et al (2007) Cdkn1c (p57Kip2) is the major regulator of embryonic growth within its imprinted domain on mouse distal chromosome BMC Dev Biol 7:53 Angiolini E, Fowden A, Coan P et al (2006) Regulation of placental efficiency for nutrient transport by imprinted genes Placenta 27:S98–S102 Anglin DM, Cohen PR, Chen H (2008) Duration of early maternal separation and prediction of schizotypal symptoms from early adolescence to midlife Schizophr Res 103:143–150 Arima T, Kamikihara T, Hayashida T et al (2005) ZAC, LIT1 (KCNQ1OT1) and p57KIP2 (CDKN1C) are in an imprinted gene network that may play a role in Beckwith-Wiedemann syndrome Nucleic Acids Res 33:2650–2660 Arking DE, Cutler DJ, Brune CW et al (2008) A common genetic variant in the neurexin superfamily member CNTNAP2 increases familial risk of autism Am J Hum Genet 82:160–164 Badcock C (2009) The imprinted brain Jessica Kingsley, London Badcock C, Crespi B (2006) Imbalanced genomic imprinting in brain development: an evolutionary basis for the aetiology of autism J Evol Biol 19:1007–1032 Baqir S, Smith LC (2006) Inhibitors of histone deacetylases and DNA methyltransferases alter imprinted gene regulation in embryonic stem cells Cloning Stem Cells 8:200–213 Barr CS, Schwandt ML, Lindell SG, Higley JD, Maestripieri D, Goldman D, Suomi SJ, Heilig M (2008) Variation at the mu-opioid receptor gene (OPRM1) influences attachment behavior in infant primates Proc Natl Acad Sci USA 105:5277–5281 Bartz JA, Hollander E (2008) Oxytocin and experimental therapeutics in autism spectrum disorders Prog Brain Res 170:451–462 Berry K, Wearden AJ, Barrowclough C, Liversidge T (2006) Attachment styles, interpersonal relationships and psychotic phenomena in a non-clinical student sample Pers Indiv Diff 41:707–718 162 B.J Crespi Berry K, Barrowclough C, Wearden A (2007) A review of the role of adult attachment style in psychosis: unexplored issues and questions for further research Clin Psychol Rev 27:458–745 Berry K, Barrowclough C, Wearden A (2008) Attachment theory: a framework for understanding symptoms and interpersonal relationships in psychosis Behav Res Ther 46:1275–1282 Blissett J, Harris G, Kirk J (2001) Feeding problems in Silver-Russell syndrome Dev Med Child Neurol 43:39–44 Bonati MT, Russo S, Finelli P et al (2007) Evaluation of autism traits in Angelman syndrome: a resource to unfold autism genes Neurogenetics 8:169–178 Bowlby J (1951) Maternal care and mental health Schocken, New York Bowlby J (1969) Attachment and loss: attachment, vol Basic Books, New York Bressan FF, De Bem TH, Perecin F et al (2009) Unearthing the roles of imprinted genes in the placenta Placenta 30:823–834 Bretherton I (1992) The origins of attachment theory: John Bowlby and Mary Ainsworth Devel Psychol 28:759–775 Bretherton I (1997) Bowlby’s legacy to developmental psychology Child Psychiatry Hum Dev 28:33–43 Brose N (2009) Synaptogenic proteins and synaptic organizers: “many hands make light work” Neuron 61:650–652 Brown WM, Consedine NS (2004) Just how happy is the happy puppet? An emotion signaling and kinship theory perspective on the behavioral phenotype of children with Angelman syndrome Med Hypotheses 63:377–385 Burt A, Trivers RL (2006) Genes in conflict: the biology of selfish genetic elements Belknap, Cambridge Champagne FA, Curley JP, Swaney WT et al (2009) Paternal influence on female behavior: the role of Peg3 in exploration, olfaction, and neuroendocrine regulation of maternal behavior of female mice Behav Neurosci 123:469–480 Charalambous M, da Rocha ST, Ferguson-Smith AC (2007) Genomic imprinting, growth control and the allocation of nutritional resources: consequences for postnatal life Curr Opin Endocrinol Diabetes Obes 14:3–12 Chetcuti A, Adams LJ, Mitchell PB, Schofield PR (2006) Altered gene expression in mice treated with the mood stabilizer sodium valproate Int J Neuropsychopharmacol 9:267–276 Cohen MM Jr (2005) Beckwith-Wiedemann syndrome: historical, clinicopathological, and etiopathogenetic perspectives Pediatr Dev Pathol 8:287–304 Cohen D, Pichard N, Tordjman S et al (2005) Specific genetic disorders and autism: clinical contribution towards their identification J Autism Dev Disord 35:103–116 Crespi B (2008a) Genomic imprinting in the development and evolution of psychotic spectrum conditions Biol Rev Camb Philos Soc 83:441–493 Crespi B (2008b) Turner syndrome and the evolution of human sexual dimorphism Evol Appl 1:449–461 Crespi B (2010) The origins and evolution of genetic disease risk in modern humans Ann N Y Acad Sci 1206:80–109 Crespi B, Badcock C (2008) Psychosis and autism as diametrical disorders of the social brain Behav Brain Sci 31:241–261, discussion 261–320 Crespi B, Summers K, Dorus S (2009) Genomic sister-disorders of neurodevelopment: an evolutionary approach Evol Appl 2:81–100 Crespi B, Stead P, Elliot M (2010) Comparative genomics of autism and schizophrenia Proc Natl Acad Sci USA 107 Suppl 1:1736–1741 Curley JP, Barton S, Surani A, Keverne EB (2004) Coadaptation in mother and infant regulated by a paternally expressed imprinted gene Proc Biol Sci 271:1303–1309 Davies W, Isles AR, Humby T, Wilkinson LS (2008a) What are imprinted genes doing in the brain? Adv Exp Med Biol 626:62–70 Davies W, Lynn PM, Relkovic D, Wilkinson LS (2008b) Imprinted genes and neuroendocrine function Front Neuroendocrinol 29:413–427 ... for social motivation and affective bonding in infants (e.g., Bowlby 1969 p 20 3? ?20 4; Moriceau and Sullivan 20 05; Grossmann et al 20 08) and children (e.g., Bartz and Hollander 20 08) have been much... sensitivity to social rejection, drug dependence, and risk of schizophrenia (Insel 20 03; Barr et al 20 08; Way et al 20 09; Mague and Blendy 20 10; Sery´ et al 20 10) and evidence for parent-of-origin effects... Distribution of parthenogenetic cells in the mouse brain and their influence on brain development and behavior Proc Natl Acad Sci USA 92: 107 82? ??10786 Andrews SC, Wood MD, Tunster SJ et al (20 07) Cdkn1c